Financial Analysis with TigerGraph

8 min readMar 24, 2022

According to CNBC, “New report finds almost 80% of active fund managers are falling behind the major indexes.” Why are they falling behind? They may have the data, they may be analyzing it but where is the disconnect?

Let’s break this down. When conducting a financial analysis and making predictions, what are those predictions based on? How one action- in the market, at a company, in macro events- affects a multitude of other actions, and how it might trigger a ripple effect. The goal is how to make money with this information. In other terms, financial analysis can be similar to what TigerGraph does for supply chain processes.

What does TigerGraph do for supply chain processes? TigerGraph helps identify potential risks, bottlenecks, the consequences from macro events and so much more. It consolidates a company’s data and eliminates the data silos by bringing data in from all sources. This allows supply chain experts to run machine learning on connected data to capitalize on potential product shortages, which can lead to reduced expenses from surcharges. Forecasting for product shortages can be similar to identifying shortages in the stock market, which can result in making money from this information.

Before getting into the details, lets make it clear- TigerGraph is a graph- but not a bar chart, pie chart, etc. Think of it like tables on steroids.

Now, to return back to how TigerGraph can be used for the financial analysis of the stock market…

Imagine being able to pull in all sorts of disparate information in one big picture. The ability to run algorithms based on unorthodox relationships in that data to find patterns to make your fund more successful.

Imagine pulling in terabytes of data from all areas of the market, writing algorithms on community detection, similarity, dependency, clustering, matching/patterns, flow, and centrality/search all using your deep financial experience.

Think of a blank canvas with unlimited opportunities to draw arrows of how one financial aspect relates to another. For example, what happens when the Fed prints more money? Inflation. What happens when inflation happens? How does it affect interest rates, stocks, bonds, precious metals, commodities, currencies? How does it create more jobs, more infrastructure, more buildings, commerce, etc.?

How would you tell the story of how you know a stock will be impacted by X or interest rates will help you make money by doing Y? How would you identify the systematic and non-systematic risks affecting the entire market? How would you use all the publicly available information to detect patterns in top companies, top performing stocks, and how that coincides with public policy decisions, inflation, the consequences it has on mortgage rates, REITs, derivatives, bonds, etc.? Then, how would you tie all of that to something so seemingly irrelevant- future weather predictions and commodities (like corn, barley, etc.)? For instance, putting in weather data over 10 years to see the patterns on farmland, predicting how it may affect barley and corn production. For example, what happens with less corn? What is corn used for? Corn can be processed into a multitude of food and industrial products including starch, sweeteners, corn oil, beverage and industrial alcohol, and fuel ethanol. What industries use those products? How will they be impacted? Would it be good to buy futures on those products?

Much of this analysis might be used with TigerGraph.

So what is TigerGraph and why does it matter?

TigerGraph is a unique kind of platform for analytics and machine learning that works on connected data.
Connected data is the key thing here. You can think of information in two parts:

What you know about things (data points) and
What you know about the relationships between those things (connections).

Other analytics platforms can only work on data points, not the connections. But that is throwing half of the information away.
In TigerGraph, you can analyze and run machine learning on both data points and connections in the same way, and this turns out to be incredibly powerful.
TigerGraph, an application that works on connected data, is the only one that can run on large, real-world datasets. TigerGraph is the only one that can operate alongside your existing data systems, working reliably in real- time.

So now that you know about what system you can use, here’s what I would recommend to get your financial imagination flowing.

Draw out on a piece a paper how one financial aspect relates to another (if you are not an expert, I’ve included a few cheat sheets at the end of this article). Map out all the information that’s publicly available about companies- their financial statements (Income Statements, Balance Sheets and Cashflow). Identify patterns in certain financial ratios to then create your own analysis of the stock’s future performance. Then tie that to the actual stock price. Use a combination of the top performing stocks in each sector to detect similarities with those other stocks you’ve just run the financial statement analysis on.

Many people may not know stocks or understand a graph database. I wanted to include a few of the algorithms’ answers you may be familiar with. Some of you may be able to understand this if you’ve seen it in a format you’re used to. I’ve included a few documents on how to build them from articles here

You can use TigerGraph to do stock predictions with time series stock similarities

here)

You can use TigerGraph to cluster stocks from various industries

and here

You can use TigerGraph to do country clustering to detect economic crises.

Similar to the above examples, with TigerGraph, you can use some algorithms to design your perfect financial portfolio of some of the following and information:

Bonds, stocks, ETFs, REITs, mutual funds, commodities, precious metals, municipal bonds, demand for credit (consumer credit, mortgage credit, business credit, government credit), supply of money, discount rate, federal funds rate, call loan rate, prime rate, Eurodollars, L.I.B.O.R, currency valuation and foreign trade, Corporate Bonds, Municipal Securities

and then use Natural Language Understanding to pull in the data from the Municipal Bond Publications and what you know about U.S. Treasury Obligations (Treasury Bills, Treasury Note, Treasury Strips, TIPS); and Agency Obligations (Federal Farm Credit System, Freddie Mac, Fannie Mae, Sallie Mae, Ginnie Mae, Derivatives), etc. Some of the datasets can be found in Yahoo Finance, Google Finance, Bloomberg, Wall Street Journal, data.world, Kaggle, CME Group and more.

Then, similar to identifying potential issues in a supply chain, use that for options contracts.

After you’ve drawn your own picture, you can build it in TigerGraph. Although you may want to break it up into various graphs, I wanted to present all the various potentials as a way to give inspiration. Below is a link to an article that can explain more of the fundamentals.

Here’s one first sketch of an example:

Figure 1: Entire Picture of how many financial instruments are related, tied to external factors and can be part of a portfolio.

Above is the entire picture but I will break it down into sections.

The first section is a way to compare multiple companies’ Financial Statements. This includes their Income Statement, Balance Sheet and Cash Flow.

Figure 2: Main areas of a how to map Companies Financial Statements

Although the details aren’t displayed, the big picture is. You’ll notice the Income Statement with Expenses, Net Earnings and Revenue, the Company Balance Sheet with Assets and Liabilities. There you can tie in the total shareholder equity and other important metrics to determine different ratios. Last, you can include the Cashflow. Also, you can tie in the External Events like important newsworthy events that you can feed in from News Information.

Figure 3: Connection to Business Fundamental Analysis

Above is how you can design your Business Fundamental Strategy and connect that to your analysis, too.

Figure 4: Pulling Actual Stock Price’s metrics

The picture above is where you can create a list of the top performing stocks in certain sectors and their actual stock price. You can then use them to analyze against all the business fundamentals and financial statements from thousands of companies (Figure 2 and Figure 3).

Here you can organize the stocks based on their industries and connect them to Figure 2- 4.

Figure 6: Financial Holdings- precious metals, bonds and commodities

Then you can tie it to your Financial Holdings which could be your stocks (Figure 5), Bonds, Commodities and Precious Metals.

Figure 7: Interest rates and their connection to bonds, precious metals and commodities

Next, you can tie how interest rates affect stocks, bonds, commodities and precious metals.

Figure 8: How External Factors like the news and weather affect stocks, commodities and precious metals

You can also use external factors like the news information and weather data to show how it will affect stock prices, commodities and precious metals. For example, you can use this information to predict futures prices and run algorithms for options. Below is an example dataset of weather data for commodities.

If you’d like to export this schema, you can do it by following the instructions below. From there, you can delete and re-arrange your schema so it aligns with your financial analysis.

Figure 9: How to import an existing solution from a shared drive.

Here is the file for the schema design and some example data. You can follow this blog to add your own vertex and edge, map your own data and then design your algorithms from our Data Science Library.

You can use our Data Science Library to start writing your own algorithms which you can find here.

Here is an example of how I mapped the data the schema.